Much of my work, I now see, has been spurred by a maddening ethnographic riddle: how to account for the vexing gap between abstractions of culture and specificities of persons and events. Initially, the problem appeared to me as the Cultural Relativity Effect. The first sentences of the first page of the preface of my first book read:
Doing anthropological fieldwork, I, like many others, sometimes experienced a frustrating relativity effect: the closer I moved toward a phenomenon, the faster it seemed to recede from my grasp. Every step forward revealed new complexities. The problem seems especially to bedevil the study of culture, something that when seen from a distance can appear monolithic and systematic but when viewed up close, in the ideas and feelings of individuals, seems to fragment into bewildering shards. (Linger 1992: vii)
Bewildering shards: a metaphor for the brute materiality and astonishing irregularity of people’s lives, so disconcertingly detached from the neat construct “culture” that pretends to speak for them. How does one reconcile such apparently antagonistic perceptions?
Obviously I am not the first writer to contemplate such questions. Repeatedly I have turned for inspiration to Edward Sapir. Early on, Sapir questioned Theodore Kroeber’s notion of culture as an evolving “superorganic” entity, divorced from human bodies and minds (Sapir 1917). Over a period of several decades he continued to warn against mistaking “fantasied universes of self-contained meaning” (1949a: 581)—that is, social-scientific abstractions—for the concrete, biographically contingent immediacies of human lives.3 Serious distortions arise when culture, viewed as a disembodied entity, is mistaken for human experience.
Unfortunately, much culture theory in the intervening years does, in spades, exactly what Sapir warned against. A big problem is the now-customary definition of culture as a system of symbols. For if, as conventionally characterized, symbols—public representations such as images, words, and rituals—are tangible forms that carry meaning, woven into a dense conceptual net, then they are in a weird sense mindlike.4 In employing the usual metaphors—symbols as “vehicles for” meanings, culture as a sticky “web”—cultural anthropologists, in the spirit of Durkheim, tend to fetishize representations. The reductio ad absurdum of this position is the idea that texts (or text-analogues) constitute an ideational cage, a cultural Supermind occupied by mindless people. Backing off from such a bizarre claim, we might more modestly propose that symbols evoke meanings in people, who draw upon their past learning and their own mental faculties in making sense of them. This premise also involves a strategic reification, as I note below, but it is a far cry from the notion that symbols constitute a weblike worldview inhabited by zombielike human beings.
The Cultural Relativity Effect appears when the notion of culture as a Supermind clashes with the empirical fact (apparent to anyone who takes a close look) that the minds supposedly dwelling within it show incredible variation, activity, and eccentricity. The effect vanishes if, as Sapir suggests, we jettison the idea of a Supermind and move culture back into human lives. This move, I argue, replaces a pseudo-problem with a set of generative questions. But I am jumping ahead. Before I get to the specifics of my proposed alternative, some general observations about anthropological theory are in order.
Geometries Versus Models
Constructing human theory resembles a mathematical exercise. Mathematicians build imaginary edifices (sets of interrelated theorems) by applying rules of inference to an inventory of axioms, which are arbitrary specifications of elementary objects and relations. Thus in mathematics, certain definitions of points and lines and their properties yield the elegant plane geometry taught in high schools; a single alteration, denial of the parallel postulate, opens the way to non-Euclidean geometries, inventions in which triangles have less or more than 180 degrees and extraordinary universes emerge. János Bolyai, a pioneer of the new geometries, wrote his father in 1823: “I have discovered things so wonderful that I was astounded . . . out of nothing I have created a strange new world” (O’Connor and Robertson 1996). Similarly, assumptions about the nature of people and groups form the base upon which one can erect elaborate, ingenious, and diverse theories of the human cosmos. And in human as in mathematical theory, a shift in the foundation can have radical effects on the superstructure.
But human theories also differ from the strange and wondrous worlds of contemporary mathematics. The new geometries created by Bolyai and others are formal systems for which internal consistency, not conformity with the world, is fundamental (Black 1959: 156–59). Such geometries are jewels of the imagination. The Greeks, in contrast, thought of geometry as the study of physical space. For all its ethereal beauty, Euclid’s geometry was a model of something else. In an important respect human theory more closely resembles ancient geometry than contemporary mathematics. Human theories are models. Models are guides and blueprints: they orient thought and action in the world. Their adequacy depends not just on coherence but also on plausibility and explanatory value. Unlike the pristine fantasies of modern geometers, theories of the human cosmos touch the earth: they abstract, and are accountable to, a reality outside themselves.
To be sure, grossly unrealistic or questionable postulates, when employed in a provisional or experimental manner, can yield theories that cast human affairs in an unfamiliar, provocative light. Such theories can provide useful elliptical accounts, and they can be good to think with, even when one knows that they are bizarre or destined eventually to fail. Significant advances in theory, however, require significant refinements in basic postulates: quantum mechanics is unimaginable without a more complex, accurate foundation than the notion of indivisible atoms advanced by Democritus. Human theory is no different. Even theories based on sophisticated, plausible postulates can, and typically (though not always) do, sooner or later collide with the world and demand reformulation.
For most of his career Sigmund Freud built his psychodynamic model around the notion that the life instincts (bodily needs, especially sexual impulses) created tensions that the human organism was compelled to discharge (a tendency Freud described as the “pleasure principle”). But after the Great War something new intruded into Freud’s model, as he indicates in this passage from his late masterpiece Civilization and Its Discontents:
men are not gentle creatures. . . . [T] hey are, on the contrary, creatures among whose instinctual endowments is to be reckoned a powerful share of aggressiveness. . . . Who, in the face of all his experience of life and of history, will have the courage to dispute this assertion? . . . Anyone who calls to mind the atrocities committed during the racial migrations or the invasions of the Huns, or by the people known as the Mongols under Jenghiz Khan and Tamerlane, or at the capture of Jerusalem by the pious Crusaders, or even, indeed, the horrors of the recent World War—anyone who calls these things to mind will have to bow humbly before the truth of this view. (1961 [1930]: 58–59)
Life instincts and the pleasure principle failed to account for the carnage of Verdun. Freud’s cumulative “experience of life and of history” forced him reluctantly to wedge the death instinct, a novel postulate, into the foundations of his psychodynamic theory. My purpose here is not to argue for or against a death instinct—though had Freud lived past 1939 he would surely have found no reason to recant. Rather, I cite Freud to suggest that, like him, every practitioner of the human sciences draws on her own compelling experience for judgments about and reformulations of premises.
Closed Theories
Premises do not always give way, because some theories, unlike Freud’s, are impervious to contrary evidence. They are more like geometries than models. All theories are, in a restricted sense, provisionally insulated from the world, because their foundational assumptions are a priori specifications; I will return to this point a bit later. But in some cases the disconnect is complete, because the assumptions permit the elaboration of a model that can explain all worldly outcomes, and so neither the model nor its assumptions come under critical scrutiny. Such self-confirming theories form perceptions rather than generating falsifiable predictions. Note that rationality—the systematic inflation of premises into models—is